From a text message in the field to a pin on the public map.
FieldworkIQ sits quietly between the reports your team receives and the Ushahidi map you publish. Below, one real report is walked through every step, in the same order it happens.
"Iko shida hapa polling station ya Mikinduri Primary. Watu wamengoja masaa matatu, BVR machine haisomi vizuri…"
"Voting machine failure at Mikinduri Primary School. Three reporters confirm. 200-person crowd, peaceful."
Someone in the field sends a tip.
An election observer at Mikinduri Primary texts your shortcode at 07:42 in the morning. The voting machine isn't reading IDs and a crowd has been waiting three hours. She writes in Swahili because she's in a hurry.
You don't need to be watching the inbox. FieldworkIQ listens to every channel your deployment is set up to receive: SMS shortcodes, your X / Twitter listening rules, a WhatsApp gateway, an inbound email address, or a web form. The report lands in your queue within seconds of being sent.
It's translated, summarised, sorted, and stripped of personal information.
Within about thirty seconds, FieldworkIQ has done the work a volunteer would normally do for the first ten minutes of an incoming report: it has translated the Swahili to English, identified the polling station by name, figured out which county and ward it's in, and matched the description to one of your deployment's categories ("Election-day disruption").
The original Swahili text is kept on file. The translation is shown to your verifier alongside it, never instead of it — so the verifier can spot mistakes and your records stay honest.
Personal information — phone numbers, names, ID numbers, any contextual identifier — is redacted before the report is staged for publishing. The original is kept in your encrypted audit trail; the public version never carries the citizen reporter's identity.
Is anyone else describing the same event?
This is the step that separates "someone said something" from "something is happening." FieldworkIQ checks every other report received in the last hour and looks for matches: same location, same kind of incident, same time window. If two or three people are describing the same event from different channels, your verifier sees that the moment they open the case.
For Mikinduri, two other people have already reported the same problem: another anonymous SMS three minutes after the first, and an X post from a long-time election observer. That's three independent sources before any volunteer has even touched it.
Everything they need, on one screen.
By the time your verifier opens the case, FieldworkIQ has assembled it. The original Swahili. The English translation. The two corroborating reports. The polling station marked on a map. The category it sorted the report into. The time it took. Whether anything sensitive was found.
Nothing is hidden. Every choice FieldworkIQ made — the translation, the category, the location — is visible, and any of them can be changed with one click before publishing. Your verifier remains the deciding voice.
Publish, edit, set aside, or escalate.
For Mikinduri, the call is clear. Three independent sources, no sensitive content, a category that matches a real election-day pattern. Your verifier reads the evidence once, glances at the corroboration, and clicks "Publish to the map." Total time on the case: about fifteen seconds.
Other cases take longer, and they should. When a report names an individual, mentions violence, or has only one source, FieldworkIQ holds it for human judgment — your verifier can edit the text, send it to a senior reviewer, or set it aside until more reporters describe the same event. Nothing is auto-published over a threshold the team didn't set.
The pin appears on Ushahidi.
Once your verifier clicks publish, the report is written to your Ushahidi deployment as a regular Post. It appears on your public map at meru-2027.uchaguzi.org within seconds, with the verified text, the category, the location, and the time it was reported. Anyone watching the live map sees the new pin.
Every step that got it there — the inbound SMS, the translation, the corroboration, the verifier's decision — is logged with timestamps. Donor reports, post-event reviews, and any regulatory inquiry can be answered from the trail FieldworkIQ keeps.
Three roles, no surprises.
FieldworkIQ is one part of a three-part workflow. Your team stays in charge of the judgment calls. Ushahidi stays your public map. FieldworkIQ handles the in-between work — the part that used to take a volunteer thirty minutes per report.
The deciders.
- Set up which channels feed the queue.
- Define your taxonomy: categories, severity thresholds, hold rules.
- Review and publish — every public post goes through a verifier.
- Escalate sensitive cases.
The in-between work.
- Receives reports from every channel you've connected.
- Translates, summarises, sorts.
- Groups reports that describe the same event.
- Holds anything sensitive for human review.
- Keeps a timestamped trail of every action.
The public map.
- Stores your verified Posts as the canonical record.
- Renders the public map your audience sees.
- Manages user accounts, categories, and deployment settings.
- Nothing changes about how Ushahidi works — FieldworkIQ just hands it cleaner reports.